Introducing the Mridula Initiative

I'm starting a small portfolio of weekend software projects about rural livelihoods in India. I'm calling it the Mridula Initiative, after my mother.

I've spent 15 years building software in healthcare, most recently leading engineering at Firefly Health. This is the work I do on the side — not for a paycheck, not for a pitch, just because it's been on my mind for a long time.

This post is the one-page summary: what it is, what it isn't, and why I'm doing it this way.

What it is

Four prototypes, each a small, deployable piece of software addressing a real problem in the rural livelihoods space:

• Mandi Price Explorer — APMC mandi prices across Uttar Pradesh districts, plotted as a choropleth with time series. Live at mandi-price-explorer.adityaksingh.dev.
• SHG Loan Rules Chatbot — A Hindi-first RAG over RBI and NABARD SHG lending guidelines. Built evaluation-first — I can show you the accuracy numbers, not just claim them.
• Organic Farm Listing App — Bilingual web app for FPOs to list produce and field buyer enquiries. Mobile-tested on a real low-end Android.
• Micro-Loan Portfolio Simulator — Model how a ₹5–10L revolving fund evolves over three years; assumptions cited from NABARD and Sa-Dhan reports.

Each project ships to a real URL with a README that's honest about the tradeoffs and at least one open question I couldn't answer from a laptop in Bengaluru. Pace: done well, not done fast.

What it isn't

It isn't a startup. It isn't a fund — yet. It isn't a pitch.

I'm not asking anyone to invest, partner, or commit to anything. This phase is for learning — about the domain, the data, the constraints — by building. Conversations with NGOs and SHGs come after, not before.

Why "Mridula"

Two reasons.

The first is the obvious one. My mother taught me to help people who don't have what I have — and that there's more to happiness than what you get from focusing on your own needs all the time. Naming the work after her is a discipline; it keeps me honest about who this is for.

The second is structural. By framing this as a personal initiative rather than a company, I get to take the slow path. No investor clock, no team to feed, no narrative to maintain. Just the work, on the weekends, until it earns the right to become something more.

The Mridula Fund

I'm including this section knowing it's premature. But naming the destination keeps the prototypes honest.

The long-term goal — somewhere in the future, after the prototypes have produced real learnings — is to seed a small revolving loan fund. Target size ₹5–10L. Source of capital: personal. Structure: to be figured out with a CA when the time comes.

What's next

Mandi Price Explorer is shipped. The SHG loan rules chatbot is next.

I'll post one note per project as each one ships, plus the occasional broader piece about what I'm learning along the way.